Hearing device comprising adaptive sound source frequency lowering

Abstract

A hearing device comprises a) an input unit for providing at least one electric input signal representing sound in a frequency sub-band representation, b) an SNR estimation unit for estimating a signal to noise ratio and/or a level estimation unit for estimating a level of said at least one electric input signal, and c) a configurable frequency transposition unit for transposing content of a source frequency sub-band FB S into a destination frequency sub-band FB D . The contents of the modified destination frequency sub-band is determined as a weighted combination of the contents of the source and destination frequency sub-bands according to the expression c1) P Dmod =αP D +βP S or c2) MAG Dmod =αMAG D +βMAG S , wherein P D and MAG D and P Dmod and MAG Dmod are the unmodified and modified power spectrum and magnitude, respectively, of the destination frequency sub-band, and P S and MAG S are the power spectrum and magnitude, respectively, of the source frequency sub-band, and the parameters α and β are destination and source band weight factors, respectively, that specify details of the frequency transposition operation. The weight factors α and β are determined in dependence of the estimate of signal to noise ratio and/or the estimate of level.

The invention claimed is: 
     
       1. A hearing device adapted to be worn at or in an ear of a user, or adapted to be fully or partially implanted in the head of the user, the hearing device comprising
 an input unit for providing at least one electric input signal representing sound in a frequency sub-band representation (k, m), where k and m are frequency and time indices, respectively, 
 an SNR estimation unit for estimating a signal to noise ratio and/or a level estimation unit for estimating a level of said at least one electric input signal, or a signal or signals derived therefrom, in said frequency sub-band representation, 
 and a configurable frequency transposition unit for transposing content of a source frequency sub-band FB S  into a destination frequency sub-band FB D  so that the contents of the resulting destination frequency sub-band is determined as a weighted combination of the contents of the source and destination frequency sub-bands according to the expression
     P   Dmod   =αP   D   +βP   S    
   or 
   MAG Dmod =αMAG D +βMAG S  
 
 
 
       wherein P D  and MAG D  are the unmodified power spectrum and magnitude, respectively, of the destination frequency sub-band before frequency transposition, P S  and MAG S  are the power spectrum and magnitude, respectively, of the source frequency sub-band, and P Dmod  and MAG Dmod  are the resulting power spectrum and magnitude, respectively, in the resulting destination sub-band after the frequency transposition, and the parameters α and β are destination and source band weight factors, respectively, that specify details of the frequency transposition operation, and wherein
 said configurable frequency transposition unit is configured to determine at least one of said weight factors α and β in dependence of said estimate of signal to noise ratio and/or said estimate of level of said at least one electric input signal, or a signal or signals derived therefrom. 
 
     
     
       2. A hearing device according to  claim 1  wherein at least one of said weight factors α and β is(are) determined in dependence of said estimate of signal to noise ratio and/or said estimate of level of said at least one electric input signal, or a signal or signals derived therefrom, in said destination and/or source frequency sub-bands. 
     
     
       3. A hearing device according to  claim 1  wherein at least one of said source and/or destination frequency sub-band(s) are/is determined in dependence of characteristics of said at least one electric input signal, or a signal or signals derived therefrom. 
     
     
       4. A hearing device according to  claim 1  wherein at least one of said source and/or destination frequency sub-band(s) is(are) pre-determined. 
     
     
       5. A hearing device according to  claim 1  configured to provide that said configurable frequency transposition unit is activated by an activation input in a specific mode of operation and/or when specific conditions are fulfilled. 
     
     
       6. A hearing device according to  claim 5  configured to provide that said specific condition comprises that the estimated signal to noise ratio (SNR) and/or the estimated level (L S ) of the source band signal is relatively high AND that the estimated signal to noise ratio (SNR) and/or the estimated level (L D ) of the destination band signal is relatively low. 
     
     
       7. A hearing device according to  claim 5  configured to provide that said specific condition is or comprises that a voice is estimated to be present in the source band signal. 
     
     
       8. A hearing device according to  claim 1  wherein said weight factors α and β are subject to a constraint. 
     
     
       9. A hearing device according to  claim 1  wherein said configurable frequency transposition unit is configured to determine said weight factors α and β under the constraint of a performance goal or a cost function. 
     
     
       10. A hearing device according to  claim 9  wherein said performance goal or cost function comprises one of a measure Î of a) listening effort b) sound quality, and c) speech intelligibility. 
     
     
       11. A hearing device according to  claim 10  wherein optimal weight factors α* and β* are determined from a database of known combinations of said weight factors (α, β), said power spectra (P D , P S ) and/or magnitudes (MAG D , MAG S ) of said destination and source frequency sub-bands, and corresponding values of the chosen measure Î. 
     
     
       12. A hearing device according to  claim 11  wherein said optimal weight factors α* and β* are determined from a database D FL  comprising corresponding values of
 P D,i , SNR(P D,i,x ), P S,j , SNR(P S,j,x ), i=1, . . . , N I , j=1, N J , x=1, . . . , N X , where N I , and N J  are the number of different values of destination and source sub-band power spectra, respectively, and N X  is the number of different values of SNR for each power spectrum value P D , P S ; 
 α* i,j,x , β* i,j,x , i=1, . . . , N I , j=1, . . . , N J , x=1, N X ; 
 
       where said optimal weight factors are determined as the values α* and β* of said weight factors α i,j,x  and β i,j,x  corresponding to a value Î(α*, β*, P D , SNR(P D ), P S , SNR(P S )) of said chosen measure Î, e.g. speech intelligibility, that fulfills, e.g. maximizes, said performance goal, or, e.g. minimizes, said cost function. 
     
     
       13. A hearing device according to  claim 1  wherein said input unit comprises a beamformer filtering unit configured to spatially filter at least two input signals representing sound in the environment said user, and providing said at least one electric input signal as a beamformed signal. 
     
     
       14. A hearing device according to  claim 1  being constituted by or comprising a hearing aid, a headset, an earphone, an ear protection device or a combination thereof. 
     
     
       15. A method of operating a hearing device, e.g. a hearing aid, adapted to be worn by a user at or in an ear of the user, or adapted to be fully or partially implanted in the head of the user, the method comprising
 providing at least one electric input signal representing sound in a frequency sub-band representation (k, m), where k and m are frequency and time indices, respectively, 
 estimating a signal to noise ratio of said at least electric input signal, or a signal or signals derived therefrom, in said time frequency representation, 
 transposing content of a source frequency sub-band FB S  into a destination frequency band FB D  so that the contents of the destination frequency sub-band is determined as a weighted combination of the contents of the source and destination frequency sub-bands according to the expression
     P   Dmod   =αP   D   +βP   S    
   or 
   MAG Dmod =αMAG D +βMAG S  
 
 
 
       wherein P D  and MAG D  is the unmodified power spectrum and magnitude, respectively, of the destination frequency sub-band before frequency transposition, P S  and MAG S  is the power spectrum or magnitude, respectively, of the source frequency sub-band, and P Dmod  and MAG Dmod  is the resulting power spectrum or magnitude, respectively, in the resulting destination sub-band after the frequency transposition, and the parameters α and β are destination and source band weight factors, respectively, that specify details of the frequency transposition operation, and
 determining at least one of said weight factors α and β in dependence of said estimate of signal to noise ratio and/or said estimate of level of said at least one electric input signal or a signal or signals derived therefrom. 
 
     
     
       16. A method according to  claim 15  comprising determining said weight factors α and β in dependence of said estimate of signal to noise ratio and/or said estimate of level of said at least one electric input signal or a signal or signals derived therefrom in said destination and/or source frequency sub-bands. 
     
     
       17. A method according to  claim 15  wherein said source frequency sub-band and said destination frequency sub-band are located on each side of a threshold frequency determined in advance of use of said hearing device with a view to a hearing profile of the user. 
     
     
       18. A method according to  claim 15  wherein said source frequency sub-band and/or said destination frequency sub-band is/are determined in advance of use of said hearing device with a view to a hearing profile of the user. 
     
     
       19. A method according to  claim 15  wherein said source frequency sub-band and/or said destination frequency sub-band is/are adaptively determined in dependence of a current electric input signal. 
     
     
       20. A non-transitory computer readable storage medium storing a computer program comprising instructions which, when the program is executed by a computer, cause the computer to carry out the method of  claim 15 .